Skip to main content

Best practices for garments with multiple styling options

When a garment can be worn in multiple ways, giving Dreem clearer visual context up front helps it generate results that match your intended styling.

Written by Phuong Anh (Sofia)

Some clothing items are harder for AI to understand from a flat lay image alone. This is especially true when the product has a specific structure, unusual shape, or multiple possible ways to wear it.

For these products, we recommend uploading images where the item is shown on a mannequin. This helps Dreem understand how the product should sit on the body, which parts should be visible, and how the garment should be styled in the final output.

When a flat lay image may not be enough

A flat lay image works well for many standard clothing items, such as t-shirts, shirts, pants, skirts, and simple dresses. However, it may be less reliable for products where the wearing method is not obvious.

Examples include:

  • Corsets

  • Ponchos

  • Shawls

  • Wraps

  • Fake sweater scarves

  • Multi-way tops or dresses

  • Layered garments

  • Cut-out or asymmetrical pieces

  • Items with straps, ties, belts, or closures

  • Accessories that can be worn in different positions

For these products, the AI may not always know whether the item should be wrapped, tied, layered, draped, fastened, worn as an accessory, or styled differently from the usual way. If you have a specific styling direction in mind, we recommend uploading the product on a mannequin so Dreem can better understand the intended look.

Why this happens

Flat lay product images show the item’s design, color, material, and shape, but they do not always show how the item should interact with the body.

For example:

  • A shawl may be worn over the shoulders, wrapped around the neck, or styled across one arm.

  • A poncho may be loose and draped, but from a flat lay it may look like a simple oversized top.

  • A corset may need to sit tightly around the waist or bust, but the AI may misread the structure if it is not shown worn.

  • A fake sweater scarf may be mistaken for a real sweater, scarf, or layered knitwear if the styling is unclear.

When the AI does not have enough visual context, the generated result may look different from your intended styling.

Recommended workflow

For clothing with a special or unclear way of wearing, use this workflow:

  1. Upload a product image on a mannequin.

    This gives the AI a clear reference for how the item should be worn.

  2. Make sure the full product is visible

    Avoid cropped images where important details such as straps, ties, sleeves, closures, or draping are hidden.

  3. Use a clear front-facing image when possible

    A clean front view helps the AI understand the main structure of the item.

  4. Add side or back views if the wearing method depends on them

    This is helpful for corsets, wrap pieces, open-back garments, layered items, or products with unusual construction.

  5. Use the most accurate styling reference available

    If the product can be worn multiple ways, upload the version that matches the result you want to generate.

Recommended input for special product types

Product type

Recommended input

Corset

On mannequin, showing how it fits around the body

Poncho

On mannequin, showing the drape and arm placement

Shawl or wrap

On mannequin, showing the intended wrapping style

Fake sweater scarf

On mannequin, showing whether it should sit as a scarf, shoulder layer, or styled knit accessory

Multi-way garment

On mannequin, showing the exact wearing option you want

Items with ties or straps

On mannequin, with ties/straps clearly visible

What to avoid

Avoid using only a flat lay image when:

  • The product can be worn in more than one way.

  • The shape looks unclear when laid flat.

  • The product depends on draping, wrapping, tying, or layering.

  • Important construction details are only understandable when worn.

  • You need the final image to follow a very specific styling direction.

A flat lay image can still be useful for showing product details, but for these special cases, it should not be the only reference image.

Summary

For standard clothing items, flat lay images are often enough. For clothing with unique, flexible, or creative ways of wearing, we recommend uploading the product on a mannequin whenever possible.

A mannequin reference helps Dreem understand how the item should fit, sit, wrap, drape, or layer on the body. This can improve the accuracy and quality of your Virtual Model, Content Kit, or Product Shot results.

This is especially helpful when the item can be styled differently from its usual use, such as a beach wrap skirt that needs to be tied in a specific way, or a scarf that should be styled as a top or shoulder layer. In these cases, a mannequin reference gives Dreem clearer visual guidance for the intended look.

Did this answer your question?